( 450 ) 



It seems worth while to call atteiilioii to the differences between 

 the diameters found i)y the same observer in ']9()() and 1901. 



I 



II 

 III 

 IV 



1900 

 0"672 ± 0"098 

 0,624 db 0,078 

 1,361 d= 0,103 

 1,277 ± 0,083 



1901 

 0",834 ± 0",()06 



,747 ± ,007 



1 ,265 ± ,009 

 1 ,169 ± ,006 



1901—1900 

 -1- 0"162 

 -f- 0,121 



— 0,096 



— 0,108 



Stone, at Oxford, once tohl me that Airy, in a conversation on 

 the determination of declinations at the meridian circle, remarked to 

 him: "I assure von. Stone, a second is a very small thing". 



If we consider the differences just adduced between the results 

 obtained by a single observer in two consecutive years we are led 

 to conclude that, for micrometer observations, even now "a tenth of 

 a second is an exceedingly small thing". 



Appendix. In how far are the tables of Damoiseau stlU reliable ? 



In the first part of this paper, pages 319 and 321, we explained 

 why we felt ourselves justified in using the tables of Damoiseau for 

 these computations in advance. We may now add that we also 

 investigated the differences of the eclipses, as observed in some recent 

 years at different observatories, from these tables, or rather from 

 the epochs given by the Nautical Almanac. In these investigations 

 we have been assisted by Mr Kress, amanuensis at the Observatory 

 of Utrecht, who has carefully searched some volumes oi' the Astroiio- 

 inische Nachrichten and of the Monthhj Notices for the time of 

 "disappearance and reappearance" of each satellite. He has further 

 combined these times, reduced them to the meridian of Greenwich, 

 and has then compared them wiih the data of the Nautical Almanac. 

 In order to simplify, we requested him to note only the observation 

 of the last light seen at disap})earance and the first light at reajipear- 

 ance^). We intended to extend our investigation from 1894 to 1905 



^) Delambre in the introduclion to his tables, does not stale explicitly the 

 precise instant to which his tables refei- but from some passages we may conclude 

 thai he also means the instant as here defined. So for instance on page LIII 

 where he says: "Les demi-dureées ont été iin pen diminuées, pour les raiyprochcr 

 des observations qiCon a faites deiniis la découverte des lunettes achromatiques" . 



That Laplace also takes it for granted that such is his real meaning, appears 

 from Gh. Vfll, 8lh book of the Mécanique Celeste. 



